1 Field survey of the 2017 Typhoon Hato and a comparison with storm 2 surge modeling in Macau 3 Linlin Li1*, Jie Yang2,3*, Chuan-Yao Lin4, Constance Ting Chua5, Yu Wang1,6, Kuifeng 4 Zhao2, Yun-Ta Wu2, Philip Li-Fan Liu2,7,8, Adam D. Switzer1,5, Kai Meng Mok9, Peitao 5 Wang10, Dongju Peng1 6 1Earth Observatory of Singapore, Nanyang Technological University, Singapore 7 2Department of Civil and Environmental Engineering, National University of Singapore, Singapore 8 3College of Harbor, Coastal and Offshore Engineering, Hohai University, China 9 4Research Center for Environmental Changes, Academia Sinica, Taipei 115, Taiwan 10 5Asian School of the Environment, Nanyang Technological University, Singapore 11 6Department of Geosciences, National Taiwan University, Taipei, Taiwan 12 7School of Civil and Environmental Engineering, Cornell University, USA 13 8Institute of Hydrological and Ocean Research, National Central University, Taiwan 14 9Department of Civil and Environmental Engineering, University of Macau, Macau, China 15 10National Marine Environmental Forecasting Center, Beijing, China 16 Corresponding to: Linlin Li ([email protected]) ; Jie Yang ([email protected]) 17 Abstract: On August 23, 2017 a Category 3 Typhoon Hato struck Southern China. Among the hardest hit cities, 18 Macau experienced the worst flooding since 1925. In this paper, we present a high-resolution survey map recording 19 inundation depths and distances at 278 sites in Macau. We show that one half of the Macau Peninsula was inundated 20 with the extent largely confined by the hilly topography. The Inner Harbor area suffered the most with the maximum 21 inundation depth of 3.1m at the coast. Using a combination of numerical models, we simulate and reproduce this 22 typhoon and storm surge event. We further investigate the effects of tidal level and sea level rise on coastal 23 inundations in Macau during the landfall of a „Hato like‟ event. 24 1 Introduction 25 On August 23, 2017, at approximately 12:50 pm local time Typhoon Hato made landfall near Zhuhai, which is 26 located on the Southern coast of Guangdong province, China (Figure 1). With an estimated 1-minute sustained wind 27 speed of 185 km/h near its center and a minimum central pressure of 945 hPa, Typhoon Hato was a Category 3 1 28 Hurricane on the Saffir-Simpson scale. Typhoon Hato was one of the strongest typhoons to affect the coastal areas 29 of the Pearl River Estuary (PRE) in Southern China over the last several decades. It caused widespread coastal 30 flooding in the PRE areas (ESCAP/WMO Typhoon Committee, 2017). Major cities in the northeast quadrant of the 31 typhoon track, including Macau, Zhuhai and Hong Kong, were severely affected. The resulting maximum storm 32 surge heights (water level above the astronomical tide) reached 1.62 m at A-Ma station in Macau, the highest since 33 water level recording began in 1925. Elsewhere in the PRE areas, a maximum storm surge of 2.79 m was recorded at 34 Zhuhai, and 1.18m, 1.65m and 2.42 m at Quarry Bay, Tai Po Kau and Tsim Bei Tsui in Hong Kong, respectively 35 (HKO, 2017) (Figure 1b). The extreme flooding in Macau was historically unprecedented in terms of the inundation 36 depth as well as the extent, and more than half of the Macau Peninsula was inundated. Typhoon Hato‟s strong wind 37 and the associated flooding resulted in 22 fatalities and caused 3.5 billion USD direct economic losses (Benfield, 38 2018). 39 Macau (and Hong Kong) commonly experience about 5-6 typhoons per year (Lee et al., 2012) and as the result the 40 low-lying area in the western part of Macau Peninsula has been frequently flooded by storm surges during major 41 typhoons. Relatively recent typhoons such as Becky (1993), Hagupit (2008), Koppu (2009), and Vicente (2012) all 42 generated storm surges that produced maximum inundation depths > 1 m in Macau, while the unnamed historical 43 typhoons in 1927 and 1948, and typhoon Gloria (1957) generated storm surges > 1.15 m (see the historical flood 44 records http://www.smg.gov.mo/smg/database/e_stormsurge_historicalRec.htm). Although frequently affected by 45 storm surges, the extreme inundation brought by Typhoon Hato still caught Macau unprepared. Consequently, the 46 local government has declared Typhoon Hato as the “worst-case” scenario and will use it as a criterion for designing 47 new engineering measures for coastal protection. 48 While Typhoon Hato has caused the worst flood event in Macau‟s history, the key flood parameters (e.g. the water 49 depth and inundation distance) have not been properly documented. Although, Macau has 2 tidal gauge stations and 50 17 inland water gauge stations distributed in the areas susceptible to flooding 51 (http://www.smg.gov.mo/smg/ftgms/e_ftgms.htm ). Unfortunately, they all failed to record the peak water level due 52 to breakdown or electricity interruption of devices during Hato (SMG, 2017). Therefore, post event surveys of key 53 flood parameters become essential for better understanding storm surge dynamics and inundation characteristics 54 (e.g. Fritz et al., 2007;Tajima et al., 2014;Takagi et al., 2017;Soria et al., 2016). For this reason, our field survey 55 team was deployed to Macau and Zhuhai on August 26, 2017 and collected flood and damage information for 5 56 days. Here, the survey data have been analyzed and used to produce a high-resolution inundation map of Macau. 57 Qualitatively speaking, several factors have contributed to the exceptional damage during Typhoon Hato: 1) 58 Typhoon Hato occurred during the second day of a Lunar month and the landfall time roughly coincided with the 59 astronomical high tide; 2) According to the record, Typhoon Hato‟s wind speed was the strongest among all the 60 typhoons in Macau since 1953. The peak wind gust reached 217.4 km/h in Taipa Grande station and broke the 61 record of 211.0 km/h set by Typhoon Ruby in 1964 (SMG, 2017;Shan et al., 2018). 3) The translation wind speed of 62 Typhoon Hato exceeded 30 km/h (Takagi et al., 2018) before its landfall, which is unusually high compared with the 63 average transitional speed of 10-15 km/h in the South China Sea (Shan et al., 2018). 4) According to Hong Kong 2 64 Observatory, Typhoon Hato momentarily became a super typhoon during its approach towards the Pearl River 65 Estuary in the morning of August 23 (HKO, 2017). The sudden intensification occurred because of the low vertical 66 wind shear and the high sea surface temperature of ~31°C in the Northern portion of the SCS (HKO, 2017). 67 It is well-known that the tide during a typhoon‟s landfall plays a significant role in the severity of the storm surge 68 induced inundation. In the case of Typhoon Hato, the coincidence of astronomical high tide and the landfall time is 69 thought to be the major factor causing the wide-spread flooding in Macau. However, using the OSU TPXO-atlas8 70 tide model (Egbert and Erofeeva, 2002) and the tide gauge location (113.551 ° E 22.167 ° N) in Macau, we 71 estimated the peak tidal level on the day of Hato‟s landfall occurred at 10:00 AM on August 23, 2017 with the tide 72 level of 0.927 m above the mean sea level (MSL), while the estimated tidal level was only 0.470 m above the MSL 73 at the reported Typhoon Hato‟s landfall time around 12:50 pm on August 23, 2017. Thus, Typhoon Hato actually 74 made landfall almost 3 hours after the peak tidal level, while the tidal level differences are almost 0.5 meter. Thus, 75 it is intriguing to ask what if Typhoon Hato had occurred at a different time with a lower or higher tidal level, how 76 would the inundation areas change? 77 To provide a quantitative answer for the question posted above, a numerical simulation tool must be validated first. 78 In this paper, the tide-surge-wave coupled hydrodynamic model, SCHISM (Semi-implicit Cross-scale Hydroscience 79 Integrated System Model) (Zhang et al., 2016) is combined with the Weather Research and Forecasting (WRF) 80 model (Skamarock et al., 2008) to simulate the entire Typhoon Hato event. The high-resolution bathymetric data in 81 PRE and topographic data in Macau are employed for calculating coastal flooding. Model-data comparisons are 82 performed to ensure that the wind fields are reproduced well by the WRF model. The field survey data (e.g. 83 inundation depth and area) are used to check the accuracy of the storm surge model. Once the numerical model is 84 validated, we can use it to conduct a series of numerical experiments to assess the possible impact of „Hato-like‟ 85 typhoon occurring at different tidal levels. Then looking at such hazard event and its counter-measures from a long- 86 term perspective, we examined the effect of sea level rise (SLR) on the inundation areas. 87 The paper is presented in the following order. We first report a high-resolution inundation map of Macau based on 88 our field measurements and observations. Then we describe each component of the numerical simulation package 89 followed by the simulation results of Typhoon Hato. Finally, we discuss the effect of tidal level and SLR through the 90 results of numerical experiments. 91 2 Post-typhoon field survey 92 On August 26, 2017, three days after Typhoon Hato made landfall, our survey team arrived at Macau, where they 93 surveyed ~ 300 sites in Macau Peninsula, measuring flow depths (water depth above the street level), maximum 94 runup, and inundation distances (Table S1).